youtube-dl/youtube_dl/extractor/nexx.py

# coding: utf-8
from __future__ import unicode_literals

import hashlib
import random
import re
import time

from .common import InfoExtractor
from ..compat import compat_str
from ..utils import (
    ExtractorError,
    int_or_none,
    parse_duration,
    try_get,
    urlencode_postdata,
)


class NexxIE(InfoExtractor):
    _VALID_URL = r'https?://api\.nexx(?:\.cloud|cdn\.com)/v3/(?P<domain_id>\d+)/videos/byid/(?P<id>\d+)'
    _TESTS = [{
        # movie
        'url': 'https://api.nexx.cloud/v3/748/videos/byid/128907',
        'md5': '16746bfc28c42049492385c989b26c4a',
        'info_dict': {
            'id': '128907',
            'ext': 'mp4',
            'title': 'Stiftung Warentest',
            'alt_title': 'Wie ein Test abläuft',
            'description': 'md5:d1ddb1ef63de721132abd38639cc2fd2',
            'release_year': 2013,
            'creator': 'SPIEGEL TV',
            'thumbnail': r're:^https?://.*\.jpg$',
            'duration': 2509,
            'timestamp': 1384264416,
            'upload_date': '20131112',
        },
        'params': {
            'format': 'bestvideo',
        },
    }, {
        # episode
        'url': 'https://api.nexx.cloud/v3/741/videos/byid/247858',
        'info_dict': {
            'id': '247858',
            'ext': 'mp4',
            'title': 'Return of the Golden Child (OV)',
            'description': 'md5:5d969537509a92b733de21bae249dc63',
            'release_year': 2017,
            'thumbnail': r're:^https?://.*\.jpg$',
            'duration': 1397,
            'timestamp': 1495033267,
            'upload_date': '20170517',
            'episode_number': 2,
            'season_number': 2,
        },
        'params': {
            'format': 'bestvideo',
            'skip_download': True,
        },
    }, {
        'url': 'https://api.nexxcdn.com/v3/748/videos/byid/128907',
        'only_matching': True,
    }]

    @staticmethod
    def _extract_urls(webpage):
        # Reference:
        # 1. https://nx-s.akamaized.net/files/201510/44.pdf

        entries = []

        # JavaScript Integration
        mobj = re.search(
            r'<script\b[^>]+\bsrc=["\']https?://require\.nexx(?:\.cloud|cdn\.com)/(?P<id>\d+)',
            webpage)
        if mobj:
            domain_id = mobj.group('id')
            for video_id in re.findall(
                    r'(?is)onPLAYReady.+?_play\.init\s*\(.+?\s*,\s*["\']?(\d+)',
                    webpage):
                entries.append(
                    'https://api.nexx.cloud/v3/%s/videos/byid/%s'
                    % (domain_id, video_id))

        # TODO: support more embed formats

        return entries

    def _handle_error(self, response):
        status = int_or_none(try_get(
            response, lambda x: x['metadata']['status']) or 200)
        if 200 <= status < 300:
            return
        raise ExtractorError(
            '%s said: %s' % (self.IE_NAME, response['metadata']['errorhint']),
            expected=True)

    def _call_api(self, domain_id, path, video_id, data=None, headers={}):
        headers['Content-Type'] = 'application/x-www-form-urlencoded; charset=UTF-8'
        result = self._download_json(
            'https://api.nexx.cloud/v3/%s/%s' % (domain_id, path), video_id,
            'Downloading %s JSON' % path, data=urlencode_postdata(data),
            headers=headers)
        self._handle_error(result)
        return result['result']

    def _real_extract(self, url):
        mobj = re.match(self._VALID_URL, url)
        domain_id, video_id = mobj.group('domain_id', 'id')

        # Reverse engineered from JS code (see getDeviceID function)
        device_id = '%d:%d:%d%d' % (
            random.randint(1, 4), int(time.time()),
            random.randint(1e4, 99999), random.randint(1, 9))

        result = self._call_api(domain_id, 'session/init', video_id, data={
            'nxp_devh': device_id,
            'nxp_userh': '',
            'precid': '0',
            'playlicense': '0',
            'screenx': '1920',
            'screeny': '1080',
            'playerversion': '6.0.00',
            'gateway': 'html5',
            'adGateway': '',
            'explicitlanguage': 'en-US',
            'addTextTemplates': '1',
            'addDomainData': '1',
            'addAdModel': '1',
        }, headers={
            'X-Request-Enable-Auth-Fallback': '1',
        })

        cid = result['general']['cid']

        # As described in [1] X-Request-Token generation algorithm is
        # as follows:
        #   md5( operation + domain_id + domain_secret )
        # where domain_secret is a static value that will be given by nexx.tv
        # as per [1]. Here is how this "secret" is generated (reversed
        # from _play.api.init function, search for clienttoken). So it's
        # actually not static and not that much of a secret.
        # 1. https://nexxtvstorage.blob.core.windows.net/files/201610/27.pdf
        secret = result['device']['clienttoken'][int(device_id[0]):]
        secret = secret[0:len(secret) - int(device_id[-1])]

        op = 'byid'

        # Reversed from JS code for _play.api.call function (search for
        # X-Request-Token)
        request_token = hashlib.md5(
            ''.join((op, domain_id, secret)).encode('utf-8')).hexdigest()

        video = self._call_api(
            domain_id, 'videos/%s/%s' % (op, video_id), video_id, data={
                'additionalfields': 'language,channel,actors,studio,licenseby,slug,subtitle,teaser,description',
                'addInteractionOptions': '1',
                'addStatusDetails': '1',
                'addStreamDetails': '1',
                'addCaptions': '1',
                'addScenes': '1',
                'addHotSpots': '1',
                'addBumpers': '1',
                'captionFormat': 'data',
            }, headers={
                'X-Request-CID': cid,
                'X-Request-Token': request_token,
            })

        general = video['general']
        title = general['title']

        stream_data = video['streamdata']
        language = general.get('language_raw') or ''

        # TODO: reverse more cdns and formats

        cdn = stream_data['cdnType']
        assert cdn == 'azure'

        azure_locator = stream_data['azureLocator']

        AZURE_URL = 'http://nx-p%02d.akamaized.net/'

        for secure in ('s', ''):
            cdn_shield = stream_data.get('cdnShieldHTTP%s' % secure.upper())
            if cdn_shield:
                azure_base = 'http%s://%s' % (secure, cdn_shield)
                break
        else:
            azure_base = AZURE_URL % int(stream_data['azureAccount'].replace('nexxplayplus', ''))

        is_ml = ',' in language
        azure_m3u8_url = '%s%s/%s_src%s.ism/Manifest(format=m3u8-aapl)' % (
            azure_base, azure_locator, video_id, ('_manifest' if is_ml else ''))

        protection_token = try_get(
            video, lambda x: x['protectiondata']['token'], compat_str)
        if protection_token:
            azure_m3u8_url += '?hdnts=%s' % protection_token

        formats = self._extract_m3u8_formats(
            azure_m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native',
            m3u8_id='%s-hls' % cdn)
        self._sort_formats(formats)

        return {
            'id': video_id,
            'title': title,
            'alt_title': general.get('subtitle'),
            'description': general.get('description'),
            'release_year': int_or_none(general.get('year')),
            'creator': general.get('studio') or general.get('studio_adref'),
            'thumbnail': try_get(
                video, lambda x: x['imagedata']['thumb'], compat_str),
            'duration': parse_duration(general.get('runtime')),
            'timestamp': int_or_none(general.get('uploaded')),
            'episode_number': int_or_none(try_get(
                video, lambda x: x['episodedata']['episode'])),
            'season_number': int_or_none(try_get(
                video, lambda x: x['episodedata']['season'])),
            'formats': formats,
        }